Share of Voice

Quick definition

Share of voice is the portion of measurable visibility a brand or domain holds in a defined market, compared with the total visibility of every domain in that same market. In SEO, the market is usually a keyword set: a fixed list of queries, in one country and language, on one device type, over a stated period.

The number is always relative. A share of voice of 18 percent says nothing until you know which keywords, which competitors, which location, and which calculation produced it. Change any of those inputs and the percentage changes, even if nothing happened in search.

Where the term comes from

The term started in advertising. There, share of voice describes one brand's media spend, or its paid presence, as a share of the total spend in a category. Advertising research also uses the idea of excess share of voice: a brand whose share of voice sits above its share of market is spending more than its size would suggest, which is studied as a possible growth signal.

SEO borrowed the name but not the input. Organic rankings are not bought, and no tool sees every query a market makes. SEO share of voice is therefore an estimate of visible presence, built from ranking data, not a record of spend. Keep that difference in mind whenever a deck places an advertising benchmark next to an organic figure.

What share of voice means in SEO

In practice, SEO share of voice answers a narrow question: across the searches we chose to watch, how much of the visible result space do we hold compared with the domains we compete against?

That question is useful because it moves the discussion from single keywords to a market segment. One keyword can jump five positions because of a SERP test. A set of 300 keywords in one product category moves more slowly and is harder to explain away. When the set is chosen well, share of voice shows whether a brand is gaining ground, holding it, or losing it to specific competitors.

It does not answer whether those searches send revenue, whether the clicks convert, or whether the brand is preferred. Those questions need analytics, conversion data, and customer research.

Three common calculations

Tools and teams use different formulas under the same label. The three below are the most common. Name the one you use in every report.

Presence share

The simplest version counts slots. Take the top 10 organic positions for every tracked keyword and count how many of them a domain holds.

positions held by the domain in the top 10 / all top-10 positions in the keyword set

Presence share is easy to explain and needs no demand data. It treats position 9 for a tiny query like position 1 for the most important query, so it says little about traffic potential.

Click-weighted visibility share

This version weights each ranking by search demand and an assumed click-through rate for its position.

sum of (search volume x expected CTR at the domain's position) / sum of the same values for all domains in scope

It is closer to traffic potential, but it adds two assumptions: the volume estimate and the CTR curve. Both are models. Real click-through varies by query, intent, brand strength, and SERP layout. Treat the result as estimated visibility, not as predicted clicks.

Share of AI answers

Some teams also count how often a brand is mentioned or cited in generated answers, for example in ChatGPT Search or Google's AI features. That is a separate measurement with its own prompt panel, run count, and coding rules. It belongs in its own column, never inside an organic share of voice percentage.

A worked example

Assume a small, hypothetical keyword set with three queries and two tracked domains, A and B. The click-through values below are illustrative assumptions for this example, not an industry benchmark: position 1 = 30 percent, position 2 = 15 percent, position 3 = 10 percent, position 4 = 6 percent.

Domain A collects 330 weighted points, domain B collects 280.

• Click-weighted share between the two tracked domains: A holds 54 percent, B holds 46 percent.
• Click-weighted share against the whole top-10 click pool: if the assumed CTR values for positions 1 to 10 add up to 75 percent, the pool is 1,700 x 0.75 = 1,275 points. A holds about 26 percent, B about 22 percent, and other domains hold the rest.
• Presence share: the set has 30 top-10 slots. A holds 2 of them, about 7 percent. B holds 3, or 10 percent.

The same rankings produce three different stories. On presence, B leads. On click-weighted share, A leads. The gap between 54 percent and 26 percent depends only on the chosen denominator. This is why a share of voice figure without its method note cannot be compared with anything.

Defining the keyword set

The keyword set is the most important decision, and it is easy to bias. Useful rules:

• Build the set from keyword research for one business segment, not from the keywords you already rank for. A set made of your own rankings will flatter you.
• Separate brand and non-brand queries. Your own brand terms inflate your share without telling you anything about the open market.
• Group keywords by intent and topic, so a gain in informational queries does not hide a loss in commercial ones.
• Include long-tail questions where they matter. AI search systems break questions into sub-queries, a process called query fan-out, so the visible market often extends beyond the short head terms.
• Freeze the set for a reporting period. If you add or remove keywords, version the set and report the break.

For larger sites, segment the set by product line, country, or funnel stage. A single sitewide percentage averages away the movements that need action.

Defining the competitor set

Business competitors and search competitors are often different. A software vendor may lose most visible positions to publishers, marketplaces, review sites, forums, and video platforms rather than to the rival named in sales calls.

Start by recording which domains actually appear in the top 10 of your set. Then decide whether to calculate share against all domains in the result pages or only against a named competitor list. Both are legitimate. A share against named competitors answers "are we beating the companies we compete with?" A share against all domains answers "how much of this search market do we actually hold?" The second number is usually much smaller and often more honest.

SERP features change the math

A position-based formula assumes ten blue links. Real result pages contain featured snippets, People Also Ask boxes, local packs, video results, shopping units, and AI Overviews. These elements shift attention and clicks, and some answers satisfy the searcher without any click at all.

A share of voice model can handle this in three ways: ignore features and say so, count feature ownership as a separate presence measure, or lower expected CTR for keywords with heavy SERP features. Whatever you choose, apply it consistently. Mixing methods between months creates movement that never happened in search.

Share of voice versus related metrics

• Market share is about sales or customers. Share of voice is about visible presence. A brand can hold a large share of voice in informational searches and still have a small market share.
• Google Ads impression share is a defined advertising metric: impressions received divided by impressions the ads were eligible for. It is not organic share of voice and the two should not be blended.
• Search Console data shows clicks, impressions, CTR, and average position for your own property only. It cannot show competitor visibility, so it cannot produce a share of voice on its own. It is, however, the best check on whether your estimated visibility turns into real impressions and clicks.
• Average position is a property-level average that ignores demand. Share of voice adds the comparison and, in the weighted version, the demand.

How to report share of voice honestly

Every share of voice figure needs a short method note. At minimum, state:

1. keyword set name, version, size, and whether brand terms are included;
2. country, language, device, and search engine;
3. competitor scope: named list or all ranking domains;
4. formula: presence share or click-weighted share, plus the CTR curve source;
5. date range and ranking check frequency;
6. how SERP features and missing rankings were handled.

Show raw counts next to percentages when the set is small. "A holds 12 of 90 top-10 slots" is easier to verify than "13 percent share of voice". When the number changes, explain which keywords moved, rather than presenting the percentage as a cause in itself.

A practical workflow

1. Choose one business segment and write down the decision the measurement should support.
2. Build and freeze the keyword set, with brand and non-brand queries tagged.
3. Record which domains rank in the top 10 and define the competitor scope.
4. Choose the formula and document the CTR assumption.
5. Track rankings on a regular schedule and calculate the share per segment.
6. Compare your estimated movement with your own Search Console clicks and impressions for the same pages.
7. Investigate the keywords behind each notable change before reporting a trend.
8. Turn losses into concrete tasks: a page to improve, a missing page to build, or a technical issue to fix.

Common mistakes

• reporting a percentage without its keyword set and formula;
• building the keyword set from your current rankings;
• mixing brand and non-brand queries in one number;
• comparing figures from two tools that use different CTR curves;
• adding AI answer mentions or ad impression share to organic share of voice;
• treating estimated visibility as measured traffic;
• changing the keyword set silently and calling the result growth;
• reacting to a single week's movement in a small set.

The Crawl Foundry perspective

Crawl Foundry does not publish a universal share of voice score, and a single score would hide the choices described above. What it provides is the raw material for a defensible calculation. You choose which keywords, lists, or segments get recurring position checks, and the rank tracker in Crawl Foundry keeps current and historical positions together with the domains visible in the top 10. Tags and landing-page context stay attached, so a loss can be traced to specific keywords and pages.

Each tracking run is paid from your balance, and the cost is shown before it starts. Search Console can be connected read-only, which gives you your own clicks and impressions as a reality check next to the estimate. The formula, the CTR assumption, and the interpretation remain your decisions, and they should be written into the report.

Related terms

• Competitive analysis
• Organic click-through rate
• SERP features
• SERP analysis
• Search volume
• Keyword difficulty
• Google Search Console
• SEO metrics
• LLM visibility
• Zero-click search

Review sources

• Wikipedia: Share of voice in advertising
• Google Search Console API: Search Analytics query method
• Google Search Central: Using Search Console and Google Analytics data for SEO
• Google Search Central: AI features and your website